Caspase-Dependent Cleavage of the Hematopoietic Specific Adaptor

Caspase-dependent cleavage of the hematopoietic speci®c adaptor protein Gads alters signalling from the T cell receptor

Donna M Berry1, Sally J Benn1, Alec M Cheng2 and C Jane McGlade*,1

1The Arthur and Sonia Labatt Brain Tumour Research Centre, The Hospital for Sick Children, Toronto, Ontario, Canada. M5G 1X8; 2Division of Rheumatology, Department of Medicine, Washington University School of Medicine, St. Louis MO, USA

Gads is a SH2 and SH3 domain-containing, hemato- apoptotic cell death (Earnshaw et al., 1999; Fadeel et poietic-speci®c adaptor protein that functions in signal- al., 2000; Villa et al., 1997). In non-apoptotic cells, ling from the T cell receptor. Gads acts by linking SLP- caspases exist as zymogens, which become active 76, bound by the carboxy-terminal Gads SH3 domain, to through either self-proteolysis or cleavage by other tyrosine phosphorylated LAT which contains binding caspase family members. Current data suggest that cell sites for the Gads SH2 domain. Gads is distinguished death signals result in the activation of caspase-8, 9 or from Grb2 and the closely related Grap protein by the 2 which then mediate cleavage and activation of presence of a 120 amino acid unique region between the downstream eector caspases (3, 6 and 7) (Earnshaw SH2 domain and the carboxy terminal SH3 domain. et al., 1999). Eector caspases are responsible for the Here we demonstrate that the unique region of Gads cleavage of speci®c target proteins, thereby mediating contains a capase cleavage site. Induction of apoptosis in the events associated with cell death. To date, a lymphocytes results in detectable Gads cleavage by number of eector caspase targets have been identi®ed 60 min. Gads cleavage is blocked in vivo by treating including structural components of the cytoskeleton cells with a caspase 3 inhibitor. A putative caspase 3 and nucleus, DNA repair enzymes and signalling cleavage site was identi®ed within the unique region and proteins involved in the regulation of cell cycle and mutation of this site prevented Gads cleavage in vitro, survival (Earnshaw et al., 1999). and in vivo. The Gads cleavage products retained the T cell receptor (TCR)-mediated apoptosis, or predicted binding speci®city for SLP-76 and LAT. activation-induced cell death (AICD) plays an Expression of the Gads cleavage products in Jurkat T important role in the negative selection of immature cells inhibited NFAT activation following TCR cross T cells within the thymus and maintaining the linking. These ®ndings indicate that cleavage of Gads in function of mature peripheral T cells (Brunner et vivo could function to alter signalling downstream of the al., 1995 Ju et al., 1995). Activation of the T cell T cell receptor by disrupting cross talk between SLP-76 receptor leads to increased transcription of both the and LAT. Oncogene (2001) 20, 1203 ± 1211. Fas receptor (CD95) and its extracellular ligand FasL (Russell, 1995). The interaction of Fas and FasL Keywords: Gads; caspase; T cell receptor; SH2 domain; initiates the cell death signalling cascade which SH3 domain mediates AICD by promoting receptor aggregation and the recruitment of the adaptor protein FADD and its associated caspase-8 (Martin et al., 1998; Introduction Muzio et al., 1998; Yang et al., 1998). Increased concentration of caspase-8 in proximity to the Apoptosis is important in the normal development and receptor leads to auto-activation of caspase-8 and homeostasis of multicellular organisms. The process of subsequent activation of downstream eector caspases apoptosis is tightly regulated by a number of and cleavage of cellular proteins responsible for biochemical signalling pathways. The Caenorhabditis controlled cellular destruction (Martin et al., 1998; elegans cell death gene ced-3 and its mammalian Muzio et al., 1998; Yan et al., 1998). homologue interleukin-1B-converting enzyme (ICE/ Initiation of signalling events from the TCR involves Caspase-1) are prototypical members of a family of phosphorylation of tyrosine residues in the intracellular cysteine-dependent aspartate directed proteases (cas- domains of the receptor CD3 complex by Src-family pases) (Nicholson et al., 1995; Yuan et al., 1993). This kinases including Lck and Fyn (Chan and Shaw, 1996; family of proteases has 13 known members which have Wange and Samelson, 1996). The phosphorylation of been shown to play an important role in mediating these residues creates a SH2 binding site for the protein tyrosine kinase ZAP-70, which is subsequently phosphorylated and activated by Lck (Chan et al., 1995; Wange et al., 1993). The activation of ZAP-70 initiates *Correspondence: CJ McGlade, The Hospital for Sick Children, 555 multiple signalling pathways through phosphorylation University Avenue, Toronto, Ontario, Canada M5G 1X8 Received 10 October 2000; revised 15 December 2000; accepted 3 of multiple downstream substrates, including enzymes January 2001 such as Phospholipase Cgl (Weiss et al., 1991), and Vav Caspase-dependent cleavage of Gads DM Berry et al 1204 (Bustelo et al., 1992), hematopoietic speci®c adaptor receptor Fas and initiates a caspase cascade. Following and docking proteins including the linker of activated treatment with FasL for up to 150 min, Jurkat cells T cells (LAT) (Buday et al., 1994; Nel et al., 1995; were harvested and protein lysates analysed by SDS ± Weber et al., 1998; Zhang et al., 1998b) and SH2 PAGE (Figure 1b). We found the appearance of anti- domain-containing leukocyte protein of 76 kDa (SLP- Gads reactive cleavage product by 90 min, at the same 76) (Jackman et al., 1995). The SLP-76 and LAT time that cleavage of PARP, a known substrate of adaptor proteins are fundamental in coordinating the caspase 3 is detected. In addition, when the Gads blot activation of both Ras-Erk and Ca2+-mobilization was stripped and reprobed for the known Gads- pathways and both are essential for T cell development associated protein SLP-76, it was discovered that it and activation (Finco et al., 1998; Jackman et al., 1995; too undergoes proteolytic cleavage after Fas receptor Yablonski et al., 1998; Zhang et al., 1998b, 1999). activation. Microscopic examination of the cells Gads is a hematopoietic adaptor protein with amino con®rmed that FasL stimulation resulted in approxi- and carboxy terminal SH3 domains ¯anking a central mately 60% of cells undergoing apoptotic cell death SH2 domain (Asada et al., 1999; Law et al., 1999; Liu after 90 min. Other agents known to trigger apoptotic and McGlade, 1998). Gads is most closely related to cell death, such as etoposide and staurosporine also the adaptors Grb2 and Grap but contains a unique resulted in cleavage of Gads in Jurkat T cells (Figure region rich in glutamine and proline residues for which 1c). Treatment of the K562 leukaemia cell line with a function has not been determined. Gads has been etoposide did not result in any observable cell death shown to play an important role in signalling from the and also did not cause cleavage of Gads. These results T cell receptor (TCR) by linking the adaptor proteins suggest that initiation of programmed cell death in T SLP-76 and LAT upon TCR engagement. SLP-76 is cells triggers proteolytic cleavage of Gads. constitutively associated with Gads through the In T cells Gads resides within speci®c protein carboxy-terminal SH3 domain of Gads, while LAT is complexes, some constitutive and some induced inducibly associated via the SH2 domain (Liu et al., following TCR stimulation. Therefore, we tested 1999). Overexpression of Gads can synergize with SLP- whether FasL stimulation also resulted in cleavage of 76 to augment NFAT and IL-2 activation, while a other components of these complexes. Following FasL SH2-inactivated form of Gads inhibits this activation stimulation for the indicated times whole cell lysates (Asada et al., 1999; Law et al., 1999; Liu et al., 1999). were prepared and immunoblotted with the indicated Here we show that in T lymphocytes, cell death anti-sera (Figure 2). Cleavage products were detected triggered by activation of the Fas receptor results in when protein lysates were blotted for the known Gads the cleavage of Gads. We show that this cleavage is associated protein LAT, and to a lesser extent the SLP- mediated by caspase 3 and that the conserved cleavage 76 associated protein SLAP-130 (Figure 2). Other LAT site lies within the unique linker region of Gads which associated molecules including Grb2, SOS and PLCg distinguishes it from Grb2 and Grap. In addition, we did not appear to be cleaved following FasL stimula- have found SLP-76 and LAT are also cleaved with tion, suggesting that Gads complexes are speci®cally similar kinetics. Due to the important role of the Gads- cleaved. SLP-76-LAT complex in TCR signalling, we also investigated the eects of the Gads caspase cleavage The linker region of Gads contains a caspase 3 products on distal signalling events. The expression of cleavage site either the C-terminal or N-terminal cleavage products inhibited NFAT activity, suggesting that caspase- Inspection of the Gads amino acid sequence revealed mediated cleavage of Gads alters signalling from the that the unique linker region between the SH2 domain activated TCR. and the carboxy terminal SH3 domain contains a potential caspase cleavage site (Figure 3a). Although the amino acid sequence throughout the linker region of human and murine Gads is not absolutely Results conserved, the sequence of the putative cleavage site at amino acids 232 ± 235 is identical in the mouse and FasL stimulation of T cells induces proteolytic cleavage human sequences (unpublished observation). Cleavage of Gads of the Gads protein at this site would yield an anti- During the course of our analysis of Gads protein Gads immunoreactive protein with a predicted mole- expression in primary lymphocytes, a faster migrating cular weight of approximately 24 kDa, in agreement protein which was immunoreactive with anti-Gads with the size of the cleavage product observed in Figure antisera was observed (Figure 1a). We considered that 1. Since the sequence in Gads most closely matched the this Gads reactive band might be generated by consensus cleavage site for the caspase 3 family of proteolytic cleavage as a result of caspase activation, proteases, we tested whether speci®c caspase inhibitors a consequence of programmed cell death which is would block Gads cleavage following FasL stimulation triggered when primary cells are placed in culture. of Jurkat T cells. Figure 3b shows that treatment of Therefore, we tested whether similar proteolytic Jurkat cells with both a pan-caspase inhibitor as well as cleavage of Gads occurred in Jurkat T cells stimulated a caspase 3 speci®c inhibitor prevented Gads cleavage. with Fas ligand (FasL), which activates the cell death An inhibitor of Caspase 1 (ICE) family members did

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1205 ab

Figure 1 Induction of apoptosis in T cells results in proteolytic cleavage of Gads and SLP-76 (a) Primary lymphocytes were isolated and cultured for 24 h. Gads was immunoprecipitated and Western blotted with anti-Gads anti-sera. The immunoreactive bands corresponding to Gads and a 26 kDa product are indicated by arrowheads. (b) Gads was immunoprecipitated from lysates of Jurkat T cells stimulated with FasL for the indicated times. The Gads immunoprecipitates were blotted with anti-Gads (upper panel). The arrows indicate the immunoreactive full length Gads and the putative cleavage product. To determine if the Gads associated molecule SLP-76 was also cleaved during apoptosis, the membrane was stripped and blotted with SLP-76 anti-sera (middle panel). The arrows indicate full length SLP-76 and a faster migrating immunoreactive species. Total cell lysates from the same FasL treated samples were blotted with anti-PARP (lower panel). Arrows indicate the full length and cleavage product of PARP. (c) Jurkat T or K562 cells were treated with etoposide or staurosporine to induce cell death. Gads was immunoprecipitated from cell lysates and membranes were blotted with anti-Gads. The arrows indicate the immunoreactive full length Gads and the putative cleavage product not protect Gads from cleavage, suggesting that Gads is also required for caspase cleavage of Gads in vivo is likely a caspase 3 target. (Figure 4b). In order to con®rm that Gads was a caspase 3 substrate and to identify the cleavage site, an in vitro Binding specificity of the Gads cleavage products cleavage assay was performed using recombinant caspases and in vitro translated 35S ± methionine Gads functions as an adaptor protein that plays a labeled Gads protein (Figure 4a). A mutant form of critical role in assembling a complex containing LAT Gads in which aspartic acid 235, within the putative and SLP-76, via its SH2 and carboxy-terminal SH3 cleavage site, was changed to alanine, was also used domain respectively, that is important for signalling in the in vitro assay. Caspase 3 eciently cleaved downstream of the T cell receptor (Liu et al., 1999). wild type Gads but not Gads D235A con®rming that Cleavage of Gads at D235 would result in the release this site is a caspase 3 cleavage site. Caspase 7 of the carboxy terminal SH3 domain of Gads from the appeared to cleave Gads very ineciently, suggesting remainder of the protein consisting of the amino that Gads is a speci®c target of caspase 3 in vitro. terminal SH3 domain and the central SH2 domain. When expressed in Jurkat T cells, only the wild-type Therefore, we examined how such a cleavage event form of Gads but not the D235A mutant was could alter the binding properties of the Gads SH2 and cleaved following FasL treatment, con®rming that SH3 domains and their interaction with targets the DIND235 sequence within the Gads linker region following T cell receptor activation. Epitope tagged

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1206 Gads proteins representative of the two cleavage TCR activation the Gads construct representing the products were expressed in Jurkat T cells. Following amino terminal cleavage product (CL-N) was found to associate with LAT while the carboxy terminal product (CL-C) did not (Figure 5). Similarly, the carboxy terminal cleavage product eciently associated with SLP-76 independent of TCR activation. Furthermore, examination of the tyrosine phosphorylated proteins bound to each cleavage product following TCR activation did not reveal the presence of any new co- immunoprecipitating proteins, suggesting the speci®city of the Gads SH2 and SH3 domains is retained even in the cleaved form of the molecule. These results suggest that cleavage of Gads could uncouple SLP-76 and LAT downstream of the TCR.

Gads cleavage products alter TCR signalling To examine the eects of Gads cleavage on downstream signalling from the T cell receptor, we measured IL-2 promoter activity in Jurkat T cells following TCR cross linking in the presence of transiently expressed Gads cleavage products. The transient over expression of SLP-76 has previously been shown to enhance transcription of a luciferase reporter gene driven by the NFAT-binding region of the IL-2 promoter in response to TCR ligation (Motto et al., 1996). Co-expression of either of the Gads caspase cleavage products with SLP- 76 inhibited this enhanced anti-CD3 dependent activa- Figure 2 Cleavage of Gads associated molecules following FasL tion of NFAT (Figure 6a). No dierence in the stimulation. Jurkat T cells were treated with FasL for the maximum luciferase units induced by ionomycin and indicated times. Protein lysates of treated cells were separated by SDS ± PAGE and blotted with the indicated anti-sera as described PMA was observed between transfectants, suggesting in Materials and methods. Arrowheads indicate the full length that the eect of the Gads cleavage products is protein and putative cleavage product in each panel proximal to the TCR (data not shown). Previously

Figure 3 The unique linker region of Gads contains a potential caspase cleavage site. (a) Schematic representation of Gads protein indicating the location of the potential caspase 3 cleavage site (DIND235) and the region of Gads recognized by the anti-Gads anti- sera. (b) Jurkat T cells were pretreated with the caspase inhibitors DEVD (caspase 3), VAD (caspase 8) or YVAD (ICE) for 30 min and then stimulated with FasL. Gads and its cleavage product were detected by immunoprecipitation and Western blotting with anti-Gads anti-sera

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1207 a

Figure 5 Gads caspase cleavage products retain binding speci- ®city. Jurkat T cells were transfected with Flag-tagged constructs corresponding to the N-terminal (CL-N) or C-terminal (CL-C) cleavage products of Gads, or Flag-tagged full-length Gads. Twenty-four hours after transfection, cells were stimulated for Figure 4 Mutation of aspartate residue 235 in the Gads linker 2 min with anti-CD3. Anti-Flag immunoprecipitations were region inhibits cleavage by caspase 3. (a) 35S-methionine labeled performed on clari®ed lysates to identify proteins bound to the wild type Gads (WT) or a mutant form of Gads in which cleavage products. Western membranes were blotted with anti- aspartate residue 235 was changed to alanine (D235A) was LAT (top panel), anti-SLP-76 (middle panel) or anti-phosphotyr- produced by in vitro translation and incubated with recombinant osine (bottom panel). The Gads cleavage products retain binding caspase 3 or caspase 7. Cleavage reactions were separated by speci®city for LAT and SLP-76. The anti-phosphotyrosine blot SDS ± PAGE and 35S-methionine labeled Gads products were did not reveal any new proteins co-precipitating with the cleavage detected by autoradiography. (b) To con®rm that D235 is also the products when compared to full length Gads. Arrows indicate caspase cleavage site in vivo, Jurkat T cells were electroporated LAT (top), SLP-76 (middle), or phospho-SLP-76 or phospho- with Flag epitope tagged wild type (WT) or mutant Gads LAT (bottom panel) (D235A) expression plasmids. Following FasL treatment Gads protein was immunoprecipitated using anti-FLAG, separated by SDS ± PAGE and membranes were blotted with anti-Gads and suggests that, in vivo, cleavage of Gads molecules could suppress signalling from the TCR. we have shown that over expression of Gads alone has no observable eect on a NFAT driven reporter, while co-expression of Gads with SLP-76 results in synergis- Discussion tic activation of NFAT (Liu et al., 1999). Expression of the Gads cleavage products was also able to inhibit the Gads is a member of a family of SH2 and SH3 synergistic activation of NFAT observed in the domain-containing adaptors which function in tyrosine presence of transient SLP-76 and wild type Gads kinase mediated signalling cascades. Gads is distin- (Figure 6a). Furthermore, when the cleavage products guished from the other two members of this family, are transiently over expressed together, a dosage Grb2 and GRAP, by the presence of a unique region dependent inhibition of SLP-76 enhanced NFAT between the central SH2 domain and the carboxy activation was observed (Figure 6b). These results terminal SH3 domain. We have shown that one indicate that the Gads cleavage products function as consequence of the presence of this unique region is dominant inhibitors of SLP-76 dependent signalling to confer sensitivity to caspase cleavage following the

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1208 a carboxy terminal SH3 domain from the rest of the Gads molecule, eectively uncoupling Gads SH3 binding proteins such as SLP-76 from the tyrosine phosphorylated proteins which bind the Gads SH2 domain. In addition, we have found that the presence of the Gads cleavage products inhibits signalling from the T cell receptor. In the periphery, reactive T cells are eliminated by activation induced cell death (AICD), a process mediated by Fas and FasL (CD95/CD95L) during the down regulation of an immune response (Brunner et al., 1995; Ju et al., 1995; Russell, 1995). Gads plays a central role in T cell activation by linking the membrane protein LAT, with SLP-76 (Liu et al., 1999). The assembly of a SLP-76-Gads-LAT complex is required for ecient transduction of downstream signals such as calcium mobilization, JNK kinase activation and NFAT activation of the IL-2 promoter (Ishiai et al., 2000; Zhang et al., 2000). Cleavage of b Gads, would result in both the loss of its adaptor function as well as the production of cleavage products which can act in a dominant inhibitory manner. Therefore, Gads cleavage would eectively result in the uncoupling of signalling pathways downstream of the T cell receptor during apoptosis and cessation of proliferation, survival and activation signals. In addition to Gads we have found that the Gads associated-molecules SLP-76 and LAT as well as SLP- 76 associated SLAP-130/Fyb are cleaved following FasL stimulation of Jurkat cells. SLP-76, LAT and SLAP-130/Fyb all contain potential caspase 3 sites which if cleaved would yield proteolytic products with molecular weights consistent with the electrophoretic mobilities of those observed in this study (unpublished observations). For example, SLP-76 contains four putative caspase 3 cleavage sites including the sequence Figure 6 Gads caspase cleavage products inhibit activation of DEDD (amino acids 320 ± 323). Cleavage at this site NFAT. (a) Gads cleavage products inhibit the enhanced NFAT would yield a 30 kDa product and a 46 kDa product activity induced by SLP-76 and Gads expression. Jurkat T cells containing the anti-SLP-76 antibody immunoreactive were transiently transfected with a reporter plasmid containing the luciferase cDNA driven by the NFAT-binding site from the region in agreement with the putative proteolytic IL2 promoter and either empty pEF-FLAG vector or FLAG- product detected in Figure 1b. Whether the physical tagged SLP-76, wild-type Gads, and/or Gads cleavage constructs association of these molecules in this complex (CL-N or CL-C). The cells were stimulated 24 h after transfection in¯uences their susceptibility to cleavage remains to with anti-CD3, anti-CD3+PMA or ionomycin+PMA (maximum stimulation) for 16 h, lysed and luciferase activity was measured be determined. with a luminometer. The results (mean of duplicates) are Other components of the T cell receptor signalling expressed as the percentage of arbitrary luciferase units induced cascade such as Vav and HPK1 have also been by ionomycin plus PMA treatment (percent of maximum). No identi®ed as caspase 3 substrates in lymphocytes (Chen dierence in the maximum luciferase units induced by ionomycin et al., 1999; Hofmann et al., 2000). Cleavage of these and PMA was observed between transfectants. (b) Gads cleavage products, when combined, inhibit the enhanced NFAT activation signalling enzymes leads to alterations in their catalytic induced by SLP-76 in a dose-dependent manner. Jurkat T cells activity. In the case of HPK1, separation of the C- were electroporated with pEF-FLAG-SLP-76 and increasing terminal regulatory domain from the N-terminal kinase amounts of the Gads CL-N and CL-C. The cells were stimulated domain leads to enhanced kinase activity, and the 24 h after transfection with anti-CD3, anti-CD3+PMA or ionomycin+PMA (maximum stimulation) for 16 h, lysed and cleaved peptides have a reduced capacity to associate luciferase activity was measured with a luminometer. The results to HPK1 binding partners Grb2 and Crk (Chen et al., (mean of duplicates) are expressed as the percentage of arbitrary 1999). The caspase-dependent cleavage of Vav1 does luciferase units induced by ionomycin plus PMA treatment (per not aect its ability to activate JNK, a kinase which cent of maximum) has been implicated as a positive regulator in apoptosis, but the cleaved form of Vav1 completely fails to induce IL-2 transcription, suggesting that the initiation of programmed cell death in T cells. Cleavage cleavage of Vav1 may inhibit signalling from the TCR of Gads at this site results in the separation of the (Hofmann et al., 2000). Gads is a unique example of an

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1209 adaptor protein that is a substrate for caspase cleavage. vector with an amino-terminal FLAG epitope tag. The The functional consequence of caspase cleavage of constructs corresponding to the putative Gads caspase Gads is likely the uncoupling of speci®c signalling cleavage products (Gads-N and Gads-C) were ampli®ed from pathways proximal to the TCR, rather than deregula- Gads full-length cDNA and cloned into pEF vector with an tion of a signalling enzyme or disassembly of a amino-terminal FLAG tag. structural complex. In addition to inducing apoptosis in T cells, there is Cell culture some evidence that Fas mediated signalling may also Jurkat cells were cultured at 378C with 5% CO2 in RPMI be involved in proliferation. Inhibition of the FADD 1640 medium supplemented with 10% (v/v) foetal bovine adaptor protein through either gene targeting or the serum (Wisent), 5 U/ml penicillin C and 5 mg/ml streptomy- expression of a dominant negative form of FADD has cin sulphate. Primary lymphocytes were isolated and cultured revealed some surprising eects on T cell proliferation. as previously described (Liu et al., 1999). Though FADD knock-out mice were embryonic lethal Apoptosis was induced by culturing cells in media with (Yeh et al., 1998; Zhang et al., 1998a), thymic 15 ng/ml of Fas ligand and 1.5 mg ml of a cross-linking reconstitution of RAG-1 de®cient chimeric mice monoclonal antibody (Upstate Biotechnology). For the caspase inhibitor experiments, cells were preincubated for revealed the expected defects in Fas-induced T cell 30 min with 50 mM AcYVAD-CHO, z-VAD-FMK, or z- apoptosis, but surprisingly cells were also resistant to DEVD-FMK (dissolved in DMSO), before stimulation with activation induced proliferation (Zhang et al., 1998a). Fas ligand. In a separate experiment, apoptosis was induced Additional evidence for FADDs proliferative role is by culturing Jurkat cells in the presence of 2 mM staurospor- provided by studies examining mice which express a ine for 3 h or 40 mM etoposide for 24 h. dominant negative form of FADD; T cells from these mice also have defects in proliferation and calcium Transient transfection and luciferase assays mobilization (Hueber et al., 2000; Zornig et al., 1998). The paradoxical eect of cell death pathways on T cell For transient transfections 26107 Jurkat cells were electro- proliferation may be mediated via the cleavage of porated with 40 mg of the indicated plasmids in electropora- tion buer (120 m KCl, 0.5 mM CaCl ,10mM K HPO , signalling molecules downstream of the T cell receptor M 2 2 4 pH 7.4, 25 mM HEPES, pH 7.5, 2 mM EGTA, pH 7.5, 5 mM which could modulate the spectrum of pathways MgC12) with a Gene Pulser (BioRad) set at 250 V and activated in response to a given signal. 960 mF. Cells were cultured for 24 h, harvested and processed as described below. For NFAT-luciferase reporter assays, 26107 Jurkat cells were electroporated with 20 mg NFAT luciferase reporter construct, together with 40 mg of empty Materials and methods pEF vector or 40 mg Gads wt, SLP-76, Gads-N or Gads-C in combinations described for each individual experiment. To Antisera and antibodies control for dierences in transfectional eciency, empty pEF Production of anity-puri®ed polyclonal Gads antibody has vector was used to be equalize the amount of DNA used in been described previously (Liu et al., 1998). The monoclonal each transfection and equivalent expression was con®rmed by anti-phosphotyrosine antibody 4G10 (Upstate Biotechnology) Western blot analysis. Twenty-four hours after transfection, was used at a dilution of 1 : 1000. Anti-FLAG M2 antibodies 56105 cells were stimulated (in duplicate) as indicated for (Sigma) were used at a dilution of 1 : 1000 for Western 16 h at 378C and lysed. Luciferase activity was quanti®ed blotting, and 5 mg were used in immunoprecipitations. with a luminometer as previously described (Fang et al., Polyclonal anti-SOS1 and anti-SOS2 antibodies were pur- 1996) and expressed in arbitrary units relative to cells treated chased from Santa Cruz Biotechnology and both were used with ionomycin and PMA (maximum stimulation). at a dilution of 1 : 500 for Western blotting. Anity-puri®ed polyclonal anti-LAT antibody (Upstate Biotechnology), anti- Immunoprecipitation and Western blotting phospholipase Cg (Upstate Biotechnology) and anti-Grb2 antibody (Transduction Laboratories) were used at a For CD3-stimulation experiments, 20 million Jurkat T cells dilutions of 1 : 250, 1 : 1000 and 1 : 500, respectively for were resuspended in 500 ml of RPMI 1640 medium, Western blotting. Anti-SLP-76 and anti-SLAP-130 polyclonal prewarmed at 378C for 10 min, and were then either left antisera (a generous gift from G Koretzky) were used at a unstimulated or were stimulated with human anti-CD3 dilution of 1 : 500 for Western blotting with a donkey anti- antibody (UCHT1-Pharmingen) for 2 min at 378C. Stimula- sheep-HRP secondary (1 : 25 000). All rabbit polyclonal tions were halted by addition of ice cold phosphatase antibodies were followed with a goat anti-rabbit-HRP from inhibitor buer (PBS pH 7.4, 10 mM Na4P2O7, 100 mM Amersham (1 : 5000) and all monoclonal antibodies were NaF, 1 mM Na3 VO4), cells were collected by centrifugation detected with sheep anti-mouse HRP (Amersham). and lysed in PLC lysis buer (50 mM HEPES pH 7.5, 150 mM NaC1, 10% glycerol, 1% Triton X-100, 1.5 mM MgC1 ,1mM EDTA, 10 mM Na ,P O , 100 mM NaF, 1 mM Plasmids and mutagenesis 2 4 2 7 Na3 VO4) containing protease inhibitors. Lysates were Wild type full-length Gads was subcloned, in frame with an clari®ed by centrifugation at 14 000 r.p.m. for 10 min at amino terminal-FLAG epitope tag, into a modi®ed pEFBOS 48C. Clari®ed lysates were then incubated for 90 min with vector containing a EF-1 promoter, as previously described protein A-sepharose or protein G-sepharose (Sigma) with (Liu et al., 1999). Gads (D235A) was made by substituting appropriate antibody as described above. Beads were then aspartic acid residue 235 with an alanine using PCR-based washed ®ve times with 1 ml NP-40 lysis buer and protein site-directed mutagenesis. Gads (D235A) was subcloned into complexes were released by boiling sepharose beads in SDS- pcDNA3.1 with an amino-terminal HA-epitope tag and pEF Laemmli sample buer for 5 min. Eluted proteins were

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1210 resolved by SDS ± PAGE and transferred to PVDF mem- reaction was then immunoprecipitated with anity-puri®ed branes for further analysis. All membranes were blocked with anti-Gads antibody for 90 min and then washed 36 with Tris-buered saline containing 1% Tween (TBST) and 5% NP-40 lysis buer. Immunoprecipitations were then incu- milk powder, or in the case of phosphotyrosine blotting, bated in 50 ml of caspase reaction buer (10 mM HEPES TBST containing 1% BSA. Following 1 h of blocking at [pH 7.4], 100 mM NaC1, 10 mM DTT, 1 mM EDTA, 0.1% room temperature, membranes were then incubated for 1 h at CHAPS) with or without 150 ng of recombinant caspase-3 or room temperature in blocking buer containing primary caspase-7 (Pharmingen). Protein samples were then separated antibody as described above. Following incubation with by SDS ± PAGE and exposed to ®lm for 2 h at room primary antibody, membranes were washed with TBST and temperature. incubated for 1 h at room temperature with appropriate secondary antibodies conjugated to horseradish peroxidase. Membranes were then washed thoroughly and antibodies Acknowledgments were visualized by enhanced chemiluminescence and auto- The authors thank Dr Stan Liu for lively discussions, Dr radiography. Gary Koretzky for anti-SLP-76 and anti-SLAP-130 anti- serum and Dr Ian Clarke for isolating primary lymphocytes. This work was funded by operating grants from the In vitro cleavage assays Cancer Research Society and from the Medical Research For in vitro transcription and translation reactions, 1 mgof Council of Canada to CJ McGlade. SJ Benn is supported pcDNA3.1 wild type or D235A Gads was incubated with by a Natural Science and Engineering Research Council rabbit reticulocyte lysate and reaction components, including Fellowship. CJ McGlade is a Research Scientist of the S35 Methionine for 90 min at 308C according to the National Cancer Institute of Canada supported with funds manufactuer's instructions (Promega). Fifteen ml of each from the Canadian Cancer Society.

References

Asada H, Ishii N, Sasaki Y, Endo K, Kasai H, Tanaka N, Liu SK, Fang N, Koretzky GA and McGlade CJ. (1999). Takeshita T, Tsuchiya S, Konno T and Sugamura K. Curr. Biol., 9, 67 ± 75. (1999). J. Exp. Med., 189, 1383 ± 1390. Liu SK and McGlade CJ. (1998). Oncogene, 17, 3073 ± 3082. Brunner T, Mogil RJ, LaFace D, Yoo NJ, Mahboubi A, Martin DA Siegel RM, Zheng L and Lenardo MJ. (1998). J. Echeverri F, Martin SJ, Force WR, Lynch DH and Ware Biol. Chem., 273, 4345 ± 4349. CF. (1995). Nature, 373, 441 ± 444. Motto DG, Ross SE, Wu J, Hendricks Taylor LR and Buday L, Egan SE, Rodriguez Viciana P, Cantrell DA and Koretzky GA. (1996). J. Exp. Med., 183, 1937 ± 1943. Downward J. (1994). J. Biol. Chem., 269, 9019 ± 9023. MuzioM,StockwellBR,StennickeHR,SalvesenGSand Bustelo XR, Ledbetter JA and Barbacid M. (1992), Nature, Dixit VM. (1998). J. Biol. Chem., 273, 2926 ± 2930. 356, 68 ± 71. Nel AE, Gupta S, Lee L, Ledbetter JA and Kanner SB. Chan AC, Dalton M, Johnson R, Kong GH, Wang T, (1995). J. Biol. Chem., 270, 18428 ± 18436. Thoma R and Kurosaki T. (1995). EMBO. J., 14, 2499 ± Nicholson DW, Ali A, Thornberry NA, Vaillancourt JP, 2508. Ding CK, Gallant M, Gareau Y, Grin PR, Labelle M Chan AC and Shaw AS. (1996). Curr. Opin. Immunol., 8, and Lazebnik YA. (1995). Nature, 376, 37 ± 43. 394 ± 401. Russell JH. (1995). Curr. Opin. Immunol., 7, 382 ± 388. Chen Y-R, Meyer CF, Ahmed B, Yao Z and Tan T-H. Villa P, Kaufmann SH and Earnshaw WC. (1997). Trends (1999). Oncogene, 18, 7370 ± 7377. Biochem. Sci., 22, 388 ± 393. Earnshaw WC, Martins LM and Kaufmann SH. (1999). Wange RL, Malek SN, Desiderio S and Samelson LE. Annu.Rev.Biochem.,68, 383 ± 424. (1993). J. Biol. Chem., 268, 19797 ± 19801. Fadeel B, Orrenius S and Zhivotovsky B. (2000). Leukemia, Wange RL and Samelson LE. (1996). Immunity, 5, 197 ± 205. 14, 1514 ± 1525. Weber JR, Orstavik S, Torgersen KM, Danbolt NC, Berg Fang N, Motto DG, Ross SE and Koretzky GA. (1996). J. SF, Ryan JC, Tasken K, Imboden JB and Vaage JT. Immunol., 157, 3769 ± 3773. (1998). J. Exp. Med., 187, 1157 ± 1161. Finco TS, Kadlecek T, Zhang W, Samelson LE and Weiss A. Weiss A, Koretzky G, Schatzman RC and Kadlecek T. (1998). Immunity, 9, 617 ± 626. (1991). Proc. Natl. Acad. Sci. USA, 88, 5484 ± 5488. Hofmann TG, Hehner SP, Droge W and Schmitz ML (2000). Yablonski D, Kuhne MR, Kadlecek T and Weiss A. (1998). Oncogene, 19, 1153 ± 1163. Science., 281, 413 ± 416. Hueber AO, Zornig M, Bernard AM, Chautan M and Evan Yang X, Chang HY and Baltimore D. (1998). Mol. Cell., 1, G. (2000). J. Biol. Chem., 275, 10453 ± 10462. 319 ± 325. Ishiai M, Kurosaki M, Inabe K, Chan AC, Sugamura K and Yeh WC, Pompa JL, McCurrach ME, Shu HB, Elia AJ, Kurosaki T. (2000). J. Exp. Med., 192, 847 ± 856. Shahinian A, Ng M, Wakeham A, Khoo W, Mitchell K, Jackman JK, Motto DG, Sun QM, Tanemoto M, Turck CW, El-Deiry WS, Lowe SW, Goeddel DV and Mak TW. Pelz GA, Koretzky GA and Findell PR. (1995). J. Biol. (1998). Science, 279, 1954 ± 1958. Chem., 270, 7029 ± 7032. Yuan J, Shaham S, Ledoux S, Ellis HM and Horvitz HR. Ju ST, Panka DJ, Cui H, Ettinger R, el-Khatib M, Sherr DH, (1993). Science, 75, 641 ± 652. Stanger BZ and Marshak-Rothstein A. (1995). Nature, Zhang J, Cado D, Chen A, Kabra NH and Winoto A 373, 444 ± 448. (1998a). Nature, 392, 296 ± 300. Law CL, Ewings MK, Chaudhary PM, Solow SA, Yun TJ, Zhang W, Sloan-Lancaster J, Kitchen J, Tible RP and Marshall AJ, Hood L and Clark EA. (1999). J. Exp. Med., Samelson LE. (1998b). Cell, 92, 83 ± 92. 189, 1243 ± 1253.

Oncogene Caspase-dependent cleavage of Gads DM Berry et al 1211 Zhang W, Sommers CL, Burshtyn DN, Stebbins CC, Zhang W, Trible RP, Zhu M, Liu SK, McGlade CJ and DeJarnette JB, Trible RP, Grinberg A, Tsay HC, Jacobs Samelson LE. (2000). J. Biol. Chem., 275, 23355 ± 23361. HM, Kessler CM, Long EO, Love PE and Samelson LE. Zornig M, Hueber AO and Evan G. (1998). Curr. Biol., 8, (1999). Immunity, 10, 323 ± 332. 467 ± 470.

Oncogene